Print version ISSN 0103-2003
On-line version ISSN 1806-9479
Rev. Econ. Sociol. Rural vol.41 no.2 Brasília Apr./June 2003
Relative intensity of bilateral trade flows, regional integration, and trade performance: the case of Brazil, 1984-1998*
Valquiria da SilvaI; Lilian C. AnefalosII; José Carlos G. dos R. FilhoIII
The objective of this research is to identify the component of trade that results specifically from bilateral relations and evaluate how the creation of trading blocs affects trade relations between countries. The trirapport coefficient of the relative intensity of bilateral agricultural sector trade flows between Brazil and other countries from 1984 and 1998 is used in the evaluation. In general, the results show that relative trade intensity between Brazil and its non-MERCOSUL trade partners fell after their entry into regional trade agreements (extra-bloc effect). The intra-bloc effect (trade expansion) is reflected by changes in trade intensity between Brazil and the other MERCOSUL members and changes in trade intensity between NAFTA members Mexico, Canada, and the United States.
Keywords: Regional Integration, Relative Intensity of Bilateral Trade, Trirapport Coefficient.
Since the 1980s, the world has seen unprecedented growth in economic regionalization, both through the creation of new agreements for integration between countries and/or zones and through the strengthening and expansion of existing blocs. One of the principal consequences of the predominance of this organizational model is an expanding role for regional agreements in international trade performance, which is the result of the net balance of positive and negative effects of integration.
Regional integration has the positive effect of creating a free trade area between signatory countries through the elimination of import barriers1, and allows competition between producers of a given good or service to take place based on productive efficiency. Another common effect is increased trade between member countries (intra-bloc trade expansion) that, when linked to the adoption of protectionist trade policy for the bloc as a whole, reduces members exposure to international competition. No less important is the increased bargaining power of countries in the bloc with other integrated countries or zones to reduce protectionism outside the bloc.
Regional trade blocs also produce negative effects. The formation of an economic bloc reinforces protectionism in relation to countries outside the bloc (extra-bloc) and generates what Viner (1950)2 calls trade diversion. The entry of a country into an existing trade bloc can also mean the adoption of new protectionist measures by this country against their old partners outside the bloc. Furthermore, when tariffs are eliminated, the country parties to the agreement lose the tariff revenue.
The socio-economic implications of the growing regionalization of global trade dynamics, and more specifically of the diversion/creation of trade, have been the subject of great interest among international economists. Their studies generally focus on (a) the performance of total trade between countries; (b) the impacts on a particular countrys economic performance, usually sectoral, after joining a free trade zone (intra-bloc effect); (c) the costs and benefits of the regional integration processes; and (d) the effects of preferential trade agreements on various countries trade performance. This paper fits into the last group of studies.
The objective of this research is to identify the component of trade that results specifically from bilateral relations and to assess how the creation of a bloc affects trade relations between countries, specifically focusing on trade between Brazil and other countries. Data used in the analysis is based on international trade patterns in the agricultural sector between 1984 and 1998. Although protectionism3 may occur in any economic sector, the principal distortions and the most complicated international regulations are found in agricultural products markets, which are generally more important to developing country economies4.
It is assumed that the reorganization of global economic space into blocs leads to the growth of trade between a blocs member countries (intra-bloc) and the expansion of each member countrys domestic market (which in effect becomes the size of the aggregate market of the member countries). As a result, a smaller slice of the blocs market is available for countries left out of the agreement.
This paper is organized in five sections including this introduction. The second section presents a brief discussion of the formation and performance of economic blocs, as well as their impacts on intra and extra-bloc relations. The third section describes the methodological approach used for this study. The fourth presents and discusses study results, and the fifth and final section presents closing remarks.
2. The influence of regional integration on international trade flows
The importance of regional agreements to the behavior of international trade is demonstrated by the fact that in 1998, the members of the European Community (EC) together with the European Free Trade Association (EFTA5), the North American Free Trade Agreement (NAFTA6) and Asia-Pacific7, a total of only 31 countries, account for an impressive 83% of all global trade; and that intra-bloc trade by members of these integrated areas has reached 67%, 36% and 51% of their respective levels of total international trade (Thorstensen, 1999).
Grether and Olarreaga (1998) have also pointed to great impact of bilateral trade agreements on international trade flows. They found preferential trade represented around 42% of total transactions in the global market in the 1993 to 1997 sub-period. Although there was only a slight (2%) total growth in preferential trade in the 1988 to 1992 sub-period, there were significant regional differences, as shown by regional bloc analyses, and much greater growth in some cases. In the EU-15, preferential trade accounted for around 72% of total trade over the entire period of their study, but was greater in the agricultural sector than in the industrial goods sector. NAFTA and MERCOSUL, the two largest trade blocs created at the end of the first sub-period, behaved differently. In NAFTA, preferential trade increased significantly in both sectors, while in MERCOSUL8, the share of total trade represented by preferential trade remained relatively stable in the largest economies but increased significantly in Paraguay and Uruguay.
Freudenberg et al. (1998) obtained similar results when they employed a gravity model to analyze the regionalization of international trade in the 1967-94 period. Using intra-zone bilateral relative intensities as an indicator, the study concluded that bilateral relative intensities of trade between the countries in Western Europe zone (EU15 + EFTA) for the most part increased: 75% for the period 1967-1994 and 81% in the sub-period 1980-1994. In the Americas, 71% of the intra-zone relative intensities improved over the entire study period, with considerable growth observed in the continents south. Only in Asia-Oceania was there a clear weakening of regional linkages.
Using a similar analytic approach, Piani and Kume (2000) also confirmed the importance of regional free trade agreements for the creation of an extraordinary level of trade between member countries, independently of whether the blocks are made up of developing or developed countries.
Carvalho (1999) analyzed the comparative performance of Brazilian agriculture in intra-MERCOSUL trade using indicators such as revealed competitiveness, the export coefficient, annual rate of trade expansion (imports and exports), and the blocs share of international trade. The author concluded that the blocs formation produced the expected effect in terms regional trade expansion.
Analyzing a performance indicator for Brazilian agricultural and agro-industrial products, Silva, Anefalos, and Reis F° (2001) confirmed that the indicators behavior was strongly influenced by the signing of regional trade agreements. From 1986 to 1998, when the creation of MERCOSUL favored Brazilian exports to the MERCOSUL member countries, Brazilian exports to NAFTA member countries decreased in approximately 60% of the product categories analyzed. Over that same period, the performance of Brazilian exportation to the European Union also deteriorated. The authors argue that this reduction in Brazilian exports to the European Union was due primarily to tariff and non-tariff barriers created by the European bloc rather than from growth in the bloc itself, which saw relatively little expansion between 1986 and 1998.
3. Methodological Approach
Basic data for the 1984-1998 period were acquired from the United Nations Statistics Division9 and from the European Department of Statistics (EUROSTAT)
The databank was structured to enable an evaluation of changes in international trade for the following agricultural product categories:
1 Live animals;
2 Meat and edible meat offal;
3 Fish, Crustaceans, Mollusks, and other aquatic invertebrates;
4 Dairy produce; birds eggs; natural honey; edible products of animal origin not elsewhere specified or included;
5 Products of animal origin not elsewhere specified or included;
6 Live trees and other; bulbs, roots and the like; cut flowers and ornamental foliage;
7 - Edible vegetables and certain roots and tubers;
8 Edible fruit and nuts; peel of citrus fruits or melons;
9 - Coffee, tea, mate and spices;
11 Milling industry products; malt; starches; insulin; wheat gluten;
12 Oil seeds and oleaginous fruits; miscellaneous grains, seeds and fruits; industrial or medical plants; straw and fodder;
13 Lacs; gums, resins and other vegetable saps and extracts;
14 Vegetable plaiting materials; vegetable products not specified or included;
15 Vegetable or animal fats and oils and their cleavage products; prepared edible fats; animal or vegetable waxes;
16 Preparations of meat, fish or crustaceans, mullusks or other aquatic invertabrates;
17- Sugars and sugar confectionery;
18 Cocoa and cocoa preparations;
19 Preparations of cereals, flour, starch or milk; pastrycooks products;
20 Preparations of vegetables, fruits, nuts or other parts of the plants;
21 Miscellaneous edible preparations;
22 Beverages, spirits and vinegar;
23 Residues and waste from de food industries; prepared animal fodder;
24 - Tobacco and manufactured tobacco substitutes;
40 Rubber and articles thereof;
41 Hides and skins (other than fur-skins) and leather;
42 Leather articles: saddles and harness; travel goods, handbags and similar containers; articles of animal gut (other than silk-worm gut);
44 Wood and articles of wood; wood charcoal;
45 Cork and articles of cork;
51 Wool, fine and coarse animal hair; yarn, and horsehair fabrics;
53 Other vegetable textile fibers; paper yarn, and fabrics woven of paper yarn.
The following countries were considered: EU15 members Germany, Austria, Spain, Finland, France, Portugal and Sweden, the signatories of NAFTA Canada, USA, and Mexico; and the members of MERCOSUL Argentina, Paraguay and Uruguay, and Brazil.
The choice of countries allowed an analysis of Brazilian trade activities with countries that had acted in isolation in the international market during some years of the studied period, as is the case of the NAFTA members, the European Economic Community members (EEC) Portugal and Spain, which entered the EU15 in 1985, and Austria, Finland and Sweden, which joined in 1995. France and Germany, which have been part of the EEC since its creation, were included to obtain the parameters of basic behavior. That is, the assumption was made that the coefficient of relative intensity of trade between Brazil and these countries has not changed significantly between 1984 and 1998.
The trirapport indicator of relative intensity, the analytical instrument used in this study, was calculated with a computational routine developed using the SAS software for Windows version 8.01. This three-dimensional indicator was used because it considers not only the country/zone of origin and the country/zone of destination but also the product categories (k).
The indicator of relative intensity is based on the relation between the observed trade flows and the theoretical flow, with the later being the expected result if the trade flows between a country i and its partner j were determined solely as a function of the size of the economy of the two partners. Thus, the indicator of relative intensity eliminates the effect of country and/or zone size in determining international trade intensity. The differences found between the coefficients obtained for theoretical trade flow and relative trade intensity correspond to the fraction of bilateral trade per category of product (trirapport) that is explained by key factors in trade creation and diversion, such as geographic proximity, trade policies, establishment of regional agreements, cultural aspects, etc. Thus, use of the Trirapport indicator allows the identification of "privileged" flow orientations by country at the global level.
The mathematical expression of the trirapport10 coefficient, defined in Freudenberg, Gaulier and Ünal-Kesenci (1998) is given by:
k, the traded product, such that,
i the country/zone of origin;
j the country/zone of destination;
Vkij the total value of transactions of product k (exports plus imports) between i and j;
Xi total exports of i, in value;
Mj total imports of j, in value;
Wk global trade of product k; and,
W global trade of all the products traded, given by
In this paper, as stated earlier, product k represents agricultural products and is the sum of the value of trade between the studied countries for the 33 selected product categories.
The trirapport indicator is composed of three distinct coefficients, which permits concurrent sectoral and geographic analysis, and indicates the intensity of change from impossibly pure and perfect competition between two trade partners that are the result of comparative advantages in exports or imports11.
Thus, if Dkij=1, only the size factor would influence bilateral trade by product category, which is not what happens in the real world. On the other hand, if Dkij > 1 or Dkij < 1 , trade flows between the two considered partners are influenced by geographic, historic, and/or political factors. Either of these two results should result in an analysis of the specific relations between the partners involved to determine the cause for the "positive" variation ( and thus an increase in the relative intensity of trade) or the "negative" variation ( , and thus a reduction in the relative intensity).
In summary, the analysis seeks to identify coherent and stable links between the different elements that make up the global trade matrix and the factors that determine these links. The factors can be regrouped into two categories (Freudenberg, Gaulier and Ünal-Kesenci, 1998):
a. Internal factors: total imports and exports within the zones studied, with the changes over time explicitly related to "internal" movements within each zone, that is, to production and export capacities, internal demand, and import requirements;
b. External Factors: a set of structural coefficients that explain the structure and intensity of bilateral trade that, while inseparable from total imports and exports, are specifically explained by external factors. These could be stable over time, with geographic, economic, cultural, or historical proximity, or subject to significant variations due to trade policy change (restrictions on imports, preferential agreements, etc.). The trirapport indicator addresses the effects of these external factors.
Finally, the trirapport coefficient, like any other coefficient used in economic analysis, is not a measure of cause and effect, but rather for indirect evaluation of phenomena observed in the real world.
4. Results and Discussion
Results show that the indicators of relative intensity of Brazilian agricultural product trade with its MERCOSUL partners were always significantly above 1, even in the years in which there was a reduction in trade. Over the entire period analyzed, these indicator values always occupied the first three positions in terms of the relative magnitude (Table 1).
Table 1 = Click to enlarge
The geographic proximity of Brazil with Argentina, Paraguay, and Uruguay certainly had an influence on this behavior, but the magnitude of the results indicates these countries participation in ALADI12 and later in MERCOSUL had a strong influence on the relative intensity of their trade with Brazil. This is indicated because the relationship among these countries geographic, cultural, and historical variables remained practically13 stable over the period under study and the size of their economies was neutralized through the calculation methodology. Thus, regional trade agreements were left as the fundamental influence on trade intensity.
Over the study period, the trirapport coefficient of Brazilian agricultural product trade with the NAFTA member countries only exceeded 1 in the case of the USA, and results show that the indicators value for trade between Brazil and the USA has dropped almost steadily since 1987, with the sharpest reduction occurring between 1988 and 1989. This behavior coincides with the 1989 signing of the Free Trade Agreement between the USA and Canada and an increase in the coefficient of relative intensity of US trade with Canada. In 1994, the year NAFTA took effect, the coefficient between Brazil and the USA recovered slightly; but since then it has declined repeatedly. In 1998, this coefficient was only 87% of the value that would be expected if economic size was the determining factor in trade between the two countries.
Brazilian trade with Canada was at its most intense at the beginning of the study period, with a relative bilateral intensity of .98 in 1984. By 1998, this indicators value had fallen to .39 (Table 1). The signing of a bilateral free trade agreement between Canada and the USA was reflected by an increase in Canadas 1990 trirapport coefficient with the USA: from 3.94 in 1988, to 4.25 in 1990. NAFTAs later implementation had an immediate effect on Canadas trade with the USA, as the Canada vs. USA trirapport coefficient rose from 4.82 in 1993, to 5.49 in 1994, and to 5.57 in 1998 (Table 2).
Table 2 = Click to enlarge
The yearly bilateral trade intensity indicator values for Brazilian trade with Mexico were always very much within what would have been expected given only the economic weights of Brazil and Mexico (Table 1). On the other hand, the NAFTA effect for Mexico, although positive for its exports of agricultural goods in the first year of the Agreement, turned neutral to slightly negative in 1996, which is an indication of Mexican economic weakness when compared with its "privileged" trading partners.
The NAFTA intra-bloc effect can be better observed by a study of trade intensity among the blocs three partners before and after the blocs creation, as given in Table 2. Analyzing the coefficients of relative intensity of bilateral trade between the USA (country i=USA) and the other countries of this bloc for the 33 products (k= 33 categories selected) between 1993 and 1994 one finds an increase from 3.99 to 4.62 for trade with Canada and from 4.91 to 5.51 for trade with Mexico.
The indicator values for Brazilian trade with European Union members Spain, Portugal, Austria, Finland, and Sweden generally decreased between 1984 and 1998. The behavior of the trirapport indicators for trade between Brazil and these countries supports the hypothesis that the bloc effect is an important determinate of trade intensity with third party (extra-bloc) countries.
The trade intensity indicator values obtained for trade between Brazil and Spain rose from 2.53 in 1984 to 2.98 in 1985 and then fell in 9 of the subsequent 13 years. The notable exceptions were between 1991 and 1992, when the indicator rose from 0.84 to 1.47, 1996 to 1997 (.57 to .67), and 1997 to 1998 (.67 to 1.0). Indicator values were significantly less than 1 from 1994 to 1997.
Indicator values for Brazilian trade with Portugal remained relatively stable and at higher levels than for Brazilian trade with Spain. Although dropping after 1987, Brazilian trade with Portugal recovered to 1984 levels in both 1993 and 1998, which may indicate that the strong cultural link with Brazil contributed to attenuate the expected effect of Portugals union in the European bloc.
Although the indicator values for Brazilian trade with Austria dropped in almost all series years, the sharpest drop, to a very unfavorable level, is found between 1994 and 199514 just after Austria joined EU15. Based on a ranking devised using the coefficients of relative intensity Austria was in eighth position as a Brazilian trade partner from 1984 to 198615, but fell to 20th in 1995 and 27th in 1998.
The intensity of Brazilian agricultural product trade with Finland stayed relatively stable over the study period. As with Austria, the greatest indicator value decline occurred from 1994 to 1995, with the coefficient dropping from 1.49 to 0.73. The indicator later recuperated to .96 in 1998. Finland improved its relative position as a Brazilian trading partner between 1985 and 1991 (from ninth to fifth) and reached its highest ranking in 1994 (fourth place).
Sweden is the final member of the group of countries that joined EU15 in 1995 to be analyzed in this study. The greatest intensity in trade between Brazil and Sweden was in the 1980s. The greatest intensity in the 1990s was in 1994, when the indicator once returned to a level greater than 1. As was observed for Austria and Finland, relative trade intensity between Brazil and Sweden fell significantly between 1994 and 1995, reaching a low that was only 64% of the value that would be expected if the economic weight of both countries were the only factor determining trade intensity. In terms of relative position in the Brazilian trading partner ranking, Sweden occupied the fifth position at the beginning of the period, fell to 11th place in 1995, and to 23rd place in 1998.
Finally, examining the performance of Brazilian trade with Germany and France, two countries with strong positions in the European bloc, one finds that the trirapport coefficient was above 1 between 1984 and 1987, then decreased in most of the next 9 years. In 1996, the coefficient had fallen to 0.32 for trade between Brazil and Germany and to 0.29 for trade between Brazil and France. The indicators value recovered considerably between 1996 and 1998, nearing 1 for trade between Brazil and Germany in 1998. This restored Germany to its position as one of Brazils top ten trade partners among the group of ranked countries.
5. Closing Remarks
In general, after eliminating the economic weight of the studied countries at the international level, there was a noticeable decline in the relative coefficients of bilateral trade intensity between Brazil and its non-MERCOSUL trading partners for the aggregated 33 categories of agricultural goods after the non-MERCOSUL countries entry into regional trade agreements that excluded Brazil, which appears to indicate an extra-bloc effect. Thus, Brazil lost space for its products in these blocs markets. This lost Brazilian trade, depending on the category, was replaced by a growth in transactions among the signatories and/or between the signatories and third party countries that, for a variety of reasons, came to receive special treatment from these blocs.
An intra-bloc effect commercial expansion can be indirectly ascertained from this studys results for Brazilian trade with its MERCOSUL partners and by the behavior of trirapport indicators between NAFTA countries. In both cases, although geographic proximity acted as a trade stimulant, clear growth in intra-bloc trade indicators is found after regional trade areas creation.
The unfavorable performances of the coefficients of relative intensity of bilateral trade between Brazil and Austria, Finland, and Sweden indicate changes in the demands that these countries imposed on Brazilian products as well as the influence of the EU15s Common Agricultural Policy on the bilateral trade of agricultural products with third parties. On the other hand, the results between Brazil and Portugal show that strong cultural links can attenuate the effect of regional agreements. This variable is not normally dominant; and as the results indicate for trade between Brazil and Argentina, the political and economic interests of traditional "rivals" can largely overcome existing cultural differences.
Finally, the Mexican experience, evaluated through agricultural product trade indicators that represent this countrys trade relations with its NAFTA partners, shows the necessity for calm, caution, and firmness in the defense of national and/or regional interests on the part of representatives of economically weak countries and/or blocs when negotiating regional agreement with more economically developed countries. This lesson is relevant for Brazil and other MERCOSUL members as they continue negotiations for the creation of the Free Trade Area of the Americas (FTAA) and the creation of an inter-regional trade area between MERCOSUL and the European Union.
BALASSA, BELA. Teoria da Integração Econômica. Lisboa: Livraria Clássica, 1964. [ Links ]
CARVALHO, Maria A. de. Comércio Agropecuário Brasileiro no MERCOSUL. Informações Econômicas, SP, v. 29, n.6, p. 7-22, jun.1999. [ Links ]
FREUDENBERG, Michael, GAULIER, Guillaume e ÜNAL-KESENCI, Deniz. La Regionalisation du Commerce International: une évaluation par les intensités relatives bilaterales. Paris: CEPII, Document de Travail n° 98-05, 1998: 103p. (extraído em 1999 de http://www.cepii.fr) [ Links ]
LAFAY, Gérard et al. Nations et Mondialisation. Paris: Economica, 1999: cap. 5. [ Links ]
PIANI, Guida e KHUME, Honorio. Fluxos Bilaterais de Comércio e Blocos Regionais: uma aplicação do modelo gravitacional. IPEA/Texto para Discussão n° 749, Rio de Janeiro, jul.2000. [ Links ]
SILVA, Valquiria, ANEFALOS, Lílian C. e REIS F°, José C. G. dos . Indicadores de Competitividade Internacional dos Produtos Agrícolas e Agroindustriais Brasileiros, 1986-1998. Agricultura em São Paulo, São Paulo, v. 48(1): 69-88, 2001. [ Links ]
THORSTENSEN, Vera. OMC: as Regras do Comércio Internacional e a Rodada do Milênio. São Paulo: Aduaneiras, 1999: 406p. [ Links ]
VINER, Jacob. The customs union issue. New York: Stevens and Sons, 1950. [ Links ]
* This study was carried out with the financial support of FAPESP through the Research Support program.
1 This was already the case in the phase of integration called Customs Union. For more on the phases of regional integration between countries see Bela Balassa, " Teoria da Integração Econômica ", Lisboa: Livraria Clássica, 1964.
2 According to Viner, regional economic integration, in eliminating customs barriers, has the effect of dislocating domestic production from one of the signatory countries to imports of the other members of the agreement, a process called trade creation. Trade deviation occurs when the socially lower-cost production of a non-member country is replaced by production of a less efficient partner for a particular good, because of the elimination of intra-bloc barriers and the creation of common external tariffs for non-member countries - which makes the final price of the non-member country artificially less competitive. Thus by maintaining extra-bloc protectionist barriers, the free trade zone could expand the market for the products of a particular member to a level greater that what would have resulted simply from the dislocation of domestic production. The extreme case could occur when the total volume of products sold within the bloc is sufficient to met its internal demand.
3 Important principally for trade deviation.
4 Analysis of the indicator of relative bilateral trade intensity for total trade flows can be found in Lafay et al. Nations et Mondialisation. Paris: Economica, 1999: chap. 5.
5 Formed by Norway, Switzerland, Iceland and Liechtenstein.
6 NAFTA consists of the United States of America, Canada and Mexico and was formed in 1994
7 Consisting of Japan, China, Korea, Hong Kong, Chinese Taipei and the Association of South East Asian Nations (ASEAN).
8 Brazil, Argentina, Uruguay and Paraguay are the signatory countries of MERCOSUL, created by the Treaty of Assunção in 1991.
9 The UN databank contains statistics on global trade for around 100 countries that present their declarations each year. According to information from the UN, the around 100 countries who submit declarations each year account for approximately 90% of the world's trade, and therefore the total obtained from these declarations about imports and exports of goods results in a very representative figure for commerce on the international market. Also, the UN does not make estimates to assess the total global sale of products.
10 The results are obtained through a more complex method than the formula presented. The complete description of construction of the trirapport indicator is found in Michael Freudenberg, Guillaume Gaulier and Deniz Ünal-Kesenci, "La Régionalisation du Commerce International: une évaluation par les intensités relatives bilaterales." CEPII, Document de Travail n° 98-05, 1998.
11 For a mathematic picture of the disaggregation of the trirapport indicator of relative intensity in three separate coefficients, including as an indicator of relative performance of exports and imports, again see Michael Freudenberg, Guillaume Gaulier and Deniz Ünal-Kesenci (1998), op.cit.
12 Latin American Integration Association, which in 1980 replaced the Latin American Free Trade Association (ALALC).
13 Practically, because in the geographic variable there is the cost of transport.
14 However, the lowest coefficient was obtained in 1996.
15 A ranking of Brazilian trade partners was developed based on the coefficients of relative intensity, and includes, along with the countries analyzed, the other members of EU15 and the "Asia-Pacific" Group, consisting of the following countries. Singapore, China, South Korea, Hong Kong, Japan, Indonesia, Malaysia and Thailand.